### Install TransSimHub with All Dependencies
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/en/installation/index.md
Install TransSimHub with all available dependencies to enable all features. This may require significant time and resources.
```bash
pip install -U -e ".[all]"
```
--------------------------------
### Fixed Installation Steps
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Corrected the installation steps in the documentation, changing `cd TransSimHub.git` to `cd TransSimHub`.
```bash
Fixed the error of visualization after installing TSHub, `Init.py -> init.py`.
```
--------------------------------
### Install TransSimHub with Documentation Features
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/en/installation/index.md
Install TransSimHub with extra dependencies for documentation editing features, such as Sphinx.
```bash
pip install -U -e ".[doc]"
```
--------------------------------
### Install TransSimHub with Pip
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/en/installation/index.md
Install TransSimHub in editable mode using pip after cloning the repository.
```bash
pip install -e .
```
--------------------------------
### Analysis Route Example
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Provides a detailed example for `analysis_route.py` within the sumotool. This script assists in the analysis and visualization of simulation results, specifically route data.
```python
from examples.sumo_tools.analysis_output.analysis_route import analysis_route
analysis_route()
```
--------------------------------
### Analysis TLS Program Example
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Presents a detailed example for analyzing traffic light programs using the sumotool. This script aids in the visualization of TLS program data.
```python
from examples.sumo_tools.analysis_output.analysis_tls_program import analysis_tls_program
analysis_tls_program()
```
--------------------------------
### Verify TransSimHub Installation
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/en/installation/index.md
Check if TransSimHub is installed correctly and print its version using Python.
```python
import tshub
print(tshub.__version__)
```
--------------------------------
### Generate Routes Example
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Demonstrates the functionality of `generate_routes.py` for creating different types of vehicles, including controlling acceleration parameters. This is a generated example.
```python
from examples.sumo_tools.generate_routes import generate_routes
generate_routes()
```
--------------------------------
### Traffic Light Choose Next Phase (Synchronize) Example
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Provides an example for the 'Choose Next Phase (Synchronize)' action in traffic light control. This action enables synchronized actions for all agents in multi-agent control tasks.
```python
from examples.traffic_light.traffic_light_action.tls_choosenextphase_syn import tls_choosenextphase_syn
tls_choosenextphase_syn()
```
--------------------------------
### Start Multi-Agent MAPPO Training
Source: https://github.com/traffic-alpha/transsimhub/blob/main/benchmark/traffic_light/README.md
Command to start the multi-agent reinforcement learning training using the MAPPO algorithm from PyTorch RL. Model weights and training logs are stored in designated folders.
```shell
python train_mappo.py
```
--------------------------------
### Start Single-Agent PPO Training
Source: https://github.com/traffic-alpha/transsimhub/blob/main/benchmark/traffic_light/README.md
Command to initiate the training process for a single-agent traffic light control system using the PPO algorithm from Stable Baselines3. Training logs and model weights are saved in specified directories.
```shell
python sb3_ppo.py
```
--------------------------------
### Vehicle Speed Scenario Setup
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Introduced the Vehicle Speed Scenario environment for controlling vehicle speed, located in the benchmark directory.
```python
Introduced [Vehicle Speed Scenario](./benchmark/sumo_envs/veh_speed/), which is accomplished by controlling vehicle speed.
```
--------------------------------
### Aircraft State Example Output
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/aircraft.md
Example of the state information returned for aircraft, including ID, type, position, speed, heading, and communication range.
```json
{
"a1": {
"id": "a1",
"aircraft_type": "drone",
"action_type": "horizontal_movement",
"position": [
1496.4644660940673,
1120.606601717798,
100
],
"speed": 5,
"heading": [
-0.7071067811865475,
0.7071067811865476,
0
],
"communication_range": 200,
"cover_radius": 173.20508075688772,
"if_sumo_visualization": true,
"img_file": "/home/wmn/TransSimHub/tshub/aircraft/./aircraft.png"
},
"a2": {
"id": "a2",
"aircraft_type": "drone",
"action_type": "horizontal_movement",
"position": [
1903.5355339059327,
796.4644660940672,
100
],
"speed": 5,
"heading": [
-0.7071067811865477,
-0.7071067811865475,
0
],
"communication_range": 200,
"cover_radius": 173.20508075688772,
"if_sumo_visualization": true,
"img_file": "/home/wmn/TransSimHub/tshub/aircraft/./aircraft.png"
}
}
```
--------------------------------
### Traffic Light Adjust Cycle Durations Example
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Shows an example of the 'Adjust Cycle Durations' action for traffic lights. This action modifies the duration of each phase within a cycle, allowing for more stable system control.
```python
from examples.traffic_light.traffic_light_action.tls_adjustCycleDuration import tls_adjustCycleDuration
tls_adjustCycleDuration()
```
--------------------------------
### Run Unit Tests with Python unittest
Source: https://github.com/traffic-alpha/transsimhub/blob/main/README.md
Execute all tests in the 'test' directory using Python's built-in unittest module. Ensure the 'tshub' package is installed and its version is greater than 1 for tests to pass.
```shell
python -m unittest discover -s test
```
--------------------------------
### Get Aircraft State Information
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/aircraft.md
Retrieve the current state information for all initialized aircraft.
```python
aircraft_state = scene_aircraft.get_objects_infos()
```
--------------------------------
### Vehicle Speed Crash Example
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Demonstrates collision detection for vehicles based on speed. This is part of the vehicle collision support features.
```python
from examples.vehicles.vehicle_action.vehicle_speed_crash import vehicle_speed_crash
vehicle_speed_crash()
```
--------------------------------
### Get Vehicle Information
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/vehicle.md
Retrieve detailed information about all vehicles in the simulation scene. This includes properties like position, speed, and emissions.
```python
data = scene_vehicles.get_objects_infos()
```
--------------------------------
### Vehicle Lane Change Crash Example
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Illustrates collision detection for vehicles during lane changes. This feature is available when the vehicle type is 'ego'.
```python
from examples.vehicles.vehicle_action.vehicle_laneChange_crash import vehicle_laneChange_crash
vehicle_laneChange_crash()
```
--------------------------------
### Build Aircraft Scene
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/aircraft.md
Initialize the aircraft builder with a SUMO connection and the defined aircraft initializations.
```python
from tshub.aircraft.aircraft_builder import AircraftBuilder
scene_aircraft = AircraftBuilder(sumo=conn, aircraft_inits=aircraft_inits)
```
--------------------------------
### Initialize Vehicle Builder
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/vehicle.md
Initialize the VehicleBuilder with a SUMO connection and specify the action type for vehicle control. The action type determines the control space available for vehicles.
```python
from tshub.vehicle.vehicle_builder import VehicleBuilder
scene_vehicles = VehicleBuilder(
sumo=conn,
action_type='lane'
)
```
--------------------------------
### 交通灯附加配置文件示例
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/sumo_tools/tls_addition.md
生成的 .add.xml 配置文件，用于在仿真中保存交通灯状态。可在 .sumocfg 文件中通过 additional-files 指定。
```xml
```
--------------------------------
### Render Scene in SUMO-GUI Mode
Source: https://github.com/traffic-alpha/transsimhub/blob/main/examples/tshub_env_render/README.md
Renders the scene in SUMO-GUI mode and saves the output directly to a specified folder. This mode requires SUMO-GUI to be enabled and the simulation window to be fullscreen for complete output.
```python
obs, reward, info, done = tshub_env.step(actions=actions)
fig = tshub_env.render(
mode='sumo_gui',
save_folder=image_save_folder
)
```
--------------------------------
### RL for Traffic Signal Control Introduction
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Provides an introduction to traffic signal control based on reinforcement learning.
```python
Added introduction to traffic signal control based on reinforcement learning, [RL for TSC](./benchmark/traffic_light/).
```
--------------------------------
### Follow Vehicle Rendering in SUMO-GUI Mode
Source: https://github.com/traffic-alpha/transsimhub/blob/main/examples/tshub_env_render/README.md
Renders the state around a specific vehicle in SUMO-GUI mode. Set focus_type to 'vehicle' and provide the vehicle's ID to focus_id. focus_distance determines the observation range.
```python
obs, reward, info, done = tshub_env.step(actions=actions)
fig = tshub_env.render(
focus_id='-1105574288#1__0__background.1',
focus_type='vehicle',
focus_distance=80,
mode='sumo_gui',
save_folder=image_save_folder
)
```
--------------------------------
### Highlight Parameter Standardization
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Moved the `highlight` parameter from `control_objects` in `vehicle_builder.py` to `init` for standardization across different objects.
```python
Moved the `highlight` parameter from `control_objects` in `vehicle_builder.py` to `init`, standardizing the `control_objects` method for different `objects`.
```
--------------------------------
### Initialize Aircraft Parameters
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/aircraft.md
Define initial parameters for aircraft, including their type, action type, position, speed, heading, and communication range.
```python
aircraft_inits = {
'a1': {
"aircraft_type": "drone",
"action_type": "horizontal_movement",
"position":(1500,1110,100), "speed":10, "heading":(1,1,0), "communication_range":200,
"if_sumo_visualization":True, "img_file":None},
'a2': {
"aircraft_type": "drone",
"action_type": "horizontal_movement",
"position":(1900,800,100), "speed":10, "heading":(1,1,0), "communication_range":200,
"if_sumo_visualization":True, "img_file":None
}
}
```
--------------------------------
### 生成交通灯附加文件
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/sumo_tools/tls_addition.md
使用 `generate_traffic_lights_additions` 函数生成自定义的 .add.xml 文件。需要指定网络文件和输出文件路径。
```python
from tshub.sumo_tools.additional_files.traffic_light_additions import generate_traffic_lights_additions
generate_traffic_lights_additions(
network_file='xxx.net.xml',
output_file='./tls.add.xml'
)
```
--------------------------------
### Global Rendering in SUMO-GUI Mode
Source: https://github.com/traffic-alpha/transsimhub/blob/main/examples/tshub_env_render/README.md
Renders the global simulation effect in SUMO-GUI mode. This is achieved by passing only the mode and save_folder parameters to the render method without specifying a focus_id.
```python
obs, reward, info, done = tshub_env.step(actions=actions)
fig = tshub_env.render(
mode='sumo_gui', # 'rgb'
save_folder=image_save_folder
)
```
--------------------------------
### Unified Connection Method
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
The connection method for `from_edge` and `direction` has been unified to the format `f"{from_edge}--{direction}"`.
```python
Unified the connection method of `from_edge` and `direction` to `f"{from_edge}--{direction}"`.
```
--------------------------------
### Execute TSC Environment Check
Source: https://github.com/traffic-alpha/transsimhub/blob/main/benchmark/traffic_light/README.md
Run this script to verify that the traffic signal control environment is set up correctly and meets expectations before proceeding with model training.
```shell
python check_tsc_env.py
```
--------------------------------
### Multi-Agent Traffic Signal Control Environment
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Introduction of the Multi-Traffic Signal Control environment, featuring 3 traffic signals and supporting multi-agent reinforcement learning algorithms like MAPPO.
```python
Introduced environment [Multi-Traffic Signal Control](./benchmark/sumo_envs/multi_junctions_tsc/), which includes $3$ traffic signals.
```
```python
Added multi-traffic signal control environment in `TsHub`, see [Multi-Agent TSC Env](./benchmark/traffic_light/multi_agents/env_utils/).
```
```python
Provided examples of `MAPPO` algorithm, controlling multiple traffic signals. Detailed algorithm can be found at [MAPPO Traffic Signal Control](./benchmark/traffic_light/multi_agents/mappo_models/).
```
--------------------------------
### Highlight Parameter Addition
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Added a `highlight` parameter in `tshub.py`.
```python
Added a `highlight` parameter in `tshub.py`.
```
--------------------------------
### Clone TransSimHub Repository
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/en/installation/index.md
Clone the official TransSimHub GitHub repository to your local machine.
```bash
git clone https://github.com/Traffic-Alpha/TransSimHub.git
cd TransSimHub
```
--------------------------------
### Local Intersection Rendering in SUMO-GUI Mode
Source: https://github.com/traffic-alpha/transsimhub/blob/main/examples/tshub_env_render/README.md
Renders a local intersection view in SUMO-GUI mode. Specify 'node' as focus_type, provide a node ID for focus_id, and set focus_distance to define the observation range.
```python
obs, reward, info, done = tshub_env.step(actions=actions)
fig = tshub_env.render(
focus_id='25663429',
focus_type='node',
focus_distance=80,
mode='sumo_gui',
save_folder=image_save_folder
)
```
--------------------------------
### Define TSC Environment Wrappers
Source: https://github.com/traffic-alpha/transsimhub/blob/main/benchmark/traffic_light/README.md
Implement state, reward, and info wrappers for traffic signal control environments. The state wrapper captures movement occupancy, and the reward wrapper calculates average queue length.
```python
def state_wrapper(self, state):
"""Returns the occupancy of each movement at the current moment.
"""
pass
def reward_wrapper(self, states) -> float:
"""Returns the average queue length at the intersection.
"""
pass
def info_wrapper(self, infos, occupancy):
"""Adds the occupancy rate of each phase to info.
"""
pass
```
--------------------------------
### Vehicle Speed Scenario Default Actions
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Updates to veh_wrapper.py to generate default actions for all vehicles, setting speed to -1 and lane to 0. Subsequent parameters only affect the ego vehicle's speed.
```python
Updated [veh_wrapper.py](./benchmark/vehicle/utils/veh_wrapper.py) with `__get_actions` and `__update_actions` methods to generate default actions for all vehicles (speed=-1, lane=0), meaning no lane changes or speed alterations. Subsequent parameters only affect the `speed` of the `ego` vehicle.
```
--------------------------------
### Vehicle Control Strategy Update
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Updated control parameters: `lane_change=-1` now uses SUMO's lane-changing strategy, and `speed=-1` uses SUMO's speed strategy. Corresponding documentation has been updated.
```python
In control, `lane_change=-1` now uses SUMO's lane-changing strategy, and `speed=-1` uses SUMO's speed strategy.
```
```python
Updated corresponding documentation.
```
--------------------------------
### Define Vehicle Types and Instances
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/vehicle.md
Define custom vehicle types with specific attributes and then create vehicle instances with routes and departure times. These are typically defined in a simulation configuration file.
```xml
```
--------------------------------
### TLS Switches 输出格式
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/sumo_tools/tls_addition.md
记录每一个 connection 绿灯的信息，包括交通灯ID、程序ID、fromLane、toLane、开始时间、结束时间和持续时间。
```xml
...
```
--------------------------------
### Traffic Light State Update
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Updated the documentation description for the new traffic light state, `fromEdge_toEdge`.
```python
Updated doc description about the new state of traffic light, `fromEdge_toEdge`.
```
--------------------------------
### Rule-Based Method Adaptation
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Updated the rule-based method in single agent to adapt to the new connection method of `from_edge` and `direction`.
```python
Updated the rule-based method in single agent to adapt to the new connection method of `from_edge` and `direction`.
```
--------------------------------
### Scenario Configuration for Special Events
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Defines parameters for a simulation scenario, including base settings, accident configurations, and special vehicle deployments. Use this to set up complex traffic events.
```python
SCENARIO_CONFIGS = {
"Hongkong_YMT_NORMAL": {
# ================== Base Scenario Parameters ==================
"SCENARIO_NAME": "Hongkong_YMT", # Scenario directory name
"SUMOCFG": "ymt_normal.sumocfg", # Combines route & network configs
"NETFILE": "./env_normal/YMT.net.xml", # Network file for map data
"JUNCTION_NAME": "J1",
"NUM_SECONDS": 500,
"PHASE_NUMBER": 3, # Number of traffic light phases
"MOVEMENT_NUMBER": 6,
"CENTER_COORDINATES": (172, 201, 60),
"SENSOR_INDEX_2_PHASE_INDEX": {0: 2, 1: 3, 2: 0, 3: 1},
# ==================== Incident Configuration ====================
"ACCIDENTS": [
{
"id": "accident_01", # Unique incident identifier
"depart_time": 20, # Simulation trigger time (seconds)
"edge_id": "30658263#0", # Target road segment ID
"lane_index": 0, # Affected lane index
"position": 99, # Location on lane (meters), lane_length-1
"duration": 50, # Duration (seconds), 0=permanent
},
{
"id": "accident_02",
"depart_time": 100,
"edge_id": "30658263#0",
"lane_index": 1,
"position": 99,
"duration": 20,
},
],
# ================== Special Vehicle Configuration ==================
"SPECIAL_VEHICLES": [
{
"id": "ambulance_01", # Unique vehicle ID
"type": "emergency", # Vehicle class
"depart_time": 10, # Departure time (sim seconds)
"route": ["960661806#0", "102640426#0"], # Path edge sequence
},
{
"id": "police_01",
"type": "police",
"depart_time": 100,
"route": ["102454134#0", "102640432#0"],
}
]
},
}
```
--------------------------------
### Control Ego Vehicles
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/vehicle.md
Continuously simulate steps, retrieve vehicle information, identify ego vehicles, generate random actions (lane change and speed), and apply these actions to control the ego vehicles.
```python
import numpy as np
while conn.simulation.getMinExpectedNumber() > 0:
# 获得车辆的信息
data = scene_vehicles.get_objects_infos()
# 控制部分车辆, 分别是 lane_change, speed
ego_vehicles = filter_ego_id(data)
actions = {_veh_id:(np.random.randint(4), None) for _veh_id in ego_vehicles}
scene_vehicles.control_objects(actions)
conn.simulationStep()
```
--------------------------------
### Plot Reward Curves Utility
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
The `plot_reward_curves.py` script is available in the utils directory for plotting reward curves with standard deviation from log files.
```python
Added [plot_reward_curves.py](./tshub/utils/plot_reward_curves.py) in utils, for plotting reward curve with standard deviation from log files.
```
--------------------------------
### TLS States 输出格式
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/sumo_tools/tls_addition.md
记录仿真每一步的交通灯状态，包括时间、ID、程序ID、相位和状态。
```xml
...
```
--------------------------------
### Vehicle Lane Change Behavior
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Fixed the issue where vehicles did not stay in the current lane when lane changes were not possible.
```python
Fixed the problem where the vehicle did not stay in the current lane when it could not change lanes [base_vehicle_action.py](./tshub/vehicle/vehicle_type/base_vehicle_action.py).
```
--------------------------------
### LaneWithContinuousSpeedAction Speed Maintenance
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Updated `LaneWithContinuousSpeedAction` to maintain the original speed when the target speed is set to -1.
```python
Updated `LaneWithContinuousSpeedAction` to maintain the original speed when the target speed is set to -1.
```
--------------------------------
### Render Scene in RGB Mode
Source: https://github.com/traffic-alpha/transsimhub/blob/main/examples/tshub_env_render/README.md
Renders the scene in RGB mode and converts the Matplotlib figure to a NumPy array using plt2arr. Requires importing plt2arr from tshub.utils.plt_to_array.
```python
from tshub.utils.plt_to_array import plt2arr
obs, reward, info, done = tshub_env.step(actions=actions)
fig = tshub_env.render(mode='rgb')
fig_array = plt2arr(fig) # convert fig to array
```
--------------------------------
### Vehicle Information Structure
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/vehicle.md
The returned vehicle information is a dictionary where keys are vehicle IDs and values are dictionaries containing detailed attributes of each vehicle.
```json
"gsndj_n7__1.0": {
"id": "gsndj_n7__1.0",
"action_type": "lane",
"vehicle_type": "car_2",
"length": 5.0,
"width": 1.8,
"heading": 307.80342137935276,
"position": [
2328.0809235224924,
523.075053998435
],
"speed": 8.775783075718211,
"road_id": "gsndj_n7",
"lane_id": "gsndj_n7_0",
"lane_index": 0,
"edges": [
"gsndj_n7",
"gsndj_n6"
],
"waiting_time": 0.0,
"accumulated_waiting_time": 0.0,
"distance": 26.36442383383401,
"co2_emission": 5280.710440252024,
"fuel_consumption": 1684.3056844797748,
"speed_without_traci": 8.775783075718211,
"leader": null,
"next_tls": [
[
"htddj_gsndj",
8,
666.575576166166,
"r"
]
]
}
```
--------------------------------
### Vehicle Lane Change Logic
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Fixed the problem of vehicle lane change direction error, where the direction was calculated based on the size of the lane index.
```python
Fixed the problem of vehicle lane change direction error, where the direction of lane change is calculated based on the size of the lane index.
```
--------------------------------
### Vehicle Speed Scenario Modifications
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Prevented direct lane changes for all vehicles to mitigate queuing at bottlenecks by adjusting speeds. Regenerated road network and traffic flow files.
```python
Prevented direct lane changes for all vehicles to mitigate queuing at bottlenecks by adjusting speeds.
```
```python
Regenerated road network and traffic flow files. Refer to [Vehicle Speed Scenario](./benchmark/sumo_envs/veh_speed/).
```
--------------------------------
### Enhanced Vehicle Environment Attributes
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
New attributes added to the vehicle environment include accumulated waiting time, distance traveled, leader information, and visualization-related attributes like width, length, and heading angle.
```python
Enhanced the `vehicle` environment with new attributes:
- `accumulated_waiting_time`: Accumulated waiting time of the vehicle.
- `distance`: Distance traveled by the vehicle.
- `leader`: Information about the vehicle ahead, including (vehicle id, distance).
- `width`, `length`, and `heading_angle`: Attributes for visualizing vehicles in the environment.
```
--------------------------------
### Control Aircraft Movement
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/aircraft.md
Issue control commands to aircraft for horizontal movement, specifying speed and heading index. Ensure 'np' is imported for random number generation.
```python
import numpy as np
actions = {
"a1": (5, np.random.randint(8)),
"a2": (5, np.random.randint(8)),
}
scene_aircraft.control_objects(actions)
```
--------------------------------
### Vehicle Feature Extraction Enhancements
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Added CO2 emissions, fuel consumption, and speed without traci to the vehicle feature extraction. Speed without traci returns the speed the vehicle would drive without speed-influencing commands.
```python
In feature extraction, added CO2 emissions (mg/s), fuel consumption (mg/s), and speed without traci (returns the speed the vehicle would drive if no speed-influencing command such as `setSpeed` or `slowDown` was given).
```
--------------------------------
### Pedestrian Module Modifications
Source: https://github.com/traffic-alpha/transsimhub/blob/main/CHANGELOG.md
Modified connection judgment in tls_connections.py to exclude pedestrian crossings and added a testing environment and state representation for pedestrians.
```python
Modified the connection judgment in [tls_connections.py](./tshub/sumo_tools/sumo_infos/tls_connections.py#76), excluding pedestrian crossings.
```
```python
Added a testing environment for pedestrians.
```
```python
Included pedestrian state representation.
```
--------------------------------
### Custom Model for Traffic Light Control
Source: https://github.com/traffic-alpha/transsimhub/blob/main/benchmark/traffic_light/README.md
Defines a custom feature extractor model using Linear and LSTM layers for embedding intersection information and encoding time-series data. This model is suitable for single-agent reinforcement learning tasks.
```python
class CustomModel(BaseFeaturesExtractor):
def __init__(self, observation_space: gym.Space, features_dim: int = 16):
"""Feature Extract
"""
super().__init__(observation_space, features_dim)
net_shape = observation_space.shape[-1] # 12
self.embedding = nn.Sequential(
nn.Linear(net_shape, 32),
nn.ReLU(),
) # 5*12 -> 5*32
self.lstm = nn.LSTM(
input_size=32, hidden_size=64,
num_layers=1, batch_first=True
)
self.relu = nn.ReLU()
self.output = nn.Sequential(
nn.Linear(64, 32),
nn.ReLU(),
nn.Linear(32, features_dim)
)
def forward(self, observations):
embedding = self.embedding(observations)
output, (hn, cn) = self.lstm(embedding)
hn = hn.view(-1, 64)
hn = self.relu(hn)
output = self.output(hn)
return output
```
--------------------------------
### Filter Ego Vehicles
Source: https://github.com/traffic-alpha/transsimhub/blob/main/docs/source/locales/zh_CN/object/vehicle.md
A utility function to filter out vehicle IDs that are designated as 'ego' vehicles from a given vehicle data dictionary. This is useful for applying specific controls only to autonomous or controlled vehicles.
```python
def filter_ego_id(vehicle_data):
ego_ids = []
for _veh_id, _veh_info in vehicle_data.items():
if _veh_info['vehicle_type'] == 'ego':
ego_ids.append(_veh_id)
return ego_ids
```
=== COMPLETE CONTENT === This response contains all available snippets from this library. No additional content exists. Do not make further requests.